Project production analysis method and device based on decision tree, and electronic equipment

Abstract

The invention discloses a project commissioning analysis method and device based on a decision tree and electronic equipment, and relates to the field of financial science and technology or other related fields, and the method comprises the steps: obtaining the commissioning data of a target project from a plurality of data sources, carrying out the preprocessing of the commissioning data, and obtaining the commissioning features, the commissioning features are input into a target decision-making tree for commissioning risk assessment, a commissioning assessment result and a commissioning success probability are output, the target decision-making tree is a decision-making tree model which is pre-constructed based on historical commissioning data and historical commissioning results and subjected to pruning processing, the contribution degree of each feature in the commissioning features is calculated based on the feature values of the commissioning features, and the commissioning risk assessment result and the commissioning success probability are output; and determining a commissioning risk feature set based on the contribution degree, and generating a commissioning risk analysis report of the target project based on the commissioning evaluation result, the commissioning success probability and the commissioning risk feature set. According to the invention, the technical problem that the accuracy is low because the risk of project commissioning depends on manual judgment in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of financial technology or other related fields, and in particular to a project production analysis method and device based on a decision tree, and an electronic device. Background Art

[0002] The launch of a project is a milestone in the life cycle of a software development project. It not only represents the end of the project development phase, but also marks the beginning of a new production phase or use phase of the project. Before a software development project is launched, it needs to go through multiple stages such as project establishment, demand analysis, software design, code development, testing, and verification. Any missing link or risk in a step may affect the software operation effect after the launch, and even cause irreparable losses to the production and operation of financial institutions.

[0003] In the related technologies, the risk analysis of software R&D project production is usually based on static analysis, which lacks the ability to comprehensively analyze complex dynamic scenarios. The existing project production analysis strategy is usually manually analyzed by project maintenance personnel with rich work experience based on their own experience. This project production analysis strategy that relies on manual analysis is inefficient and prone to errors. In the case of complex R&D processes and the participation of multiple parties, it can often only qualitatively analyze the risks of project production, and cannot quantitatively predict the probability of risk occurrence and the consequences of project production.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a project commissioning analysis method and device, and electronic equipment based on a decision tree, so as to at least solve the technical problem in the related art that the risk of project commissioning is judged manually with low accuracy.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a project commissioning analysis method based on a decision tree is provided, comprising: obtaining commissioning data of a target project from multiple data sources, preprocessing the commissioning data, and obtaining commissioning characteristics, wherein the commissioning characteristics include: categorical characteristics and numerical characteristics; inputting the commissioning characteristics into a target decision tree, performing a commissioning risk assessment by the target decision tree, and outputting a commissioning assessment result and a commissioning success probability, wherein the target decision tree is a decision tree model that is pre-constructed based on historical commissioning data and historical commissioning assessment results and has undergone pruning processing; based on the characteristic values ​​of the commissioning characteristics, calculating the contribution of each feature in the commissioning characteristics, and determining a set of commissioning risk characteristics based on the contribution; and generating a commissioning risk analysis report for the target project based on the commissioning assessment results, the commissioning success probability, and the set of commissioning risk characteristics.

[0007] Optionally, the step of preprocessing the production data to obtain production characteristics includes: obtaining all production data, wherein the production data includes at least one of the following: production type, development team level, development cycle days, production team level, requirement complexity, resource availability, code scanning results, test coverage, and number of version defects, wherein the code scanning results include: number of code lines and number of code files; classifying the production type, development team level, production team level, requirement complexity, and resource availability to obtain classified characteristics; normalizing the development cycle days, number of code lines, number of code files, test coverage, and number of version defects to obtain numerical characteristics.

[0008] Optionally, before pre-building a target decision tree based on historical production data and historical production evaluation results, the process includes: obtaining historical production data of historical projects and corresponding historical production evaluation results, pre-processing the historical production data to obtain historical production characteristics; and dividing the historical production characteristics and the corresponding historical production evaluation results into a training set and a test set.

[0009] Optionally, the step of pre-constructing a target decision tree based on historical production data and historical production evaluation results includes: constructing a target decision tree based on historical production features and historical production evaluation results in a training set, and the step of constructing the target decision tree includes: performing a correlation check on each historical production feature and the corresponding historical production evaluation result, and determining the historical production feature whose verification result indicates a first correlation as a leaf node of the decision tree; wherein, when the number of categories of the classified historical production features is greater than a preset classification number threshold, before performing the correlation check, performing multi-classification feature processing on the classified historical production features; excluding the historical production features that have been determined as leaf nodes, determining the next leaf node from the remaining historical production features, and iteratively determining the next leaf node until the decision tree meets the preset growth stop condition; and performing post-pruning processing on the decision tree that has stopped growing to obtain the target decision tree.

[0010] Optionally, the step of performing multi-classification feature processing on the classified historical production features includes: calculating the information gain of each category in the classified historical production features, and sorting the information gains; combining the categories corresponding to the first-ranked and second-ranked information gains into target features of the classified historical production features, and using the target features for correlation verification.

[0011] Optionally, the step of pre-constructing a target decision tree based on historical production data and historical production evaluation results also includes: testing the target decision tree constructed by the training set based on the historical production features and historical production evaluation results in the test set, and the step of testing the target decision tree includes: inputting the historical production data into the target decision tree to obtain the test set production evaluation results and the test set production success probability; based on the historical production evaluation results and the test set production evaluation results, calculating the accuracy, precision and recall of the target decision tree; based on the accuracy, precision and recall, adjusting the parameters of the target decision tree, and using the adjusted target decision tree for production risk analysis of the target project.

[0012] Optionally, the step of determining a set of production risk characteristics based on contribution includes: sorting the absolute values ​​of contribution; determining the production characteristics corresponding to the first-ranked contribution as the first production risk characteristic; determining the production characteristics corresponding to other contribution degrees excluding the first-ranked contribution as the second production risk characteristic; and generating a set of production risk characteristics based on the first production risk characteristic and the second production risk characteristic.

[0013] According to another aspect of an embodiment of the present invention, there is also provided a project commissioning analysis device based on a decision tree, comprising: an acquisition unit, used to acquire commissioning data of a target project from multiple data sources, pre-process the commissioning data, and obtain commissioning characteristics, wherein the commissioning characteristics include: categorical characteristics and numerical characteristics; an evaluation unit, used to input the commissioning characteristics into a target decision tree, perform commissioning risk evaluation by the target decision tree, and output a commissioning evaluation result and a commissioning success probability, wherein the target decision tree is a decision tree model that is pre-constructed based on historical commissioning data and historical commissioning evaluation results and has undergone pruning processing; a calculation unit, based on the characteristic value of the commissioning characteristics, calculates the contribution of each feature in the commissioning characteristics, and determines a set of commissioning risk characteristics based on the contribution; a generation unit, based on the commissioning evaluation results, the probability of commissioning success, and the set of commissioning risk characteristics, generates a commissioning risk analysis report for the target project.

[0014] Optionally, the acquisition unit includes: a first acquisition module, used to acquire all production data when pre-processing the production data to obtain production features, wherein the production data includes at least one of the following: production type, development team level, development cycle days, production team level, requirement complexity, resource availability, code scanning results, test coverage, and number of version defects, wherein the code scanning results include: number of code lines and number of code files; a classification module, used to classify the production type, development team level, production team level, requirement complexity and resource availability to obtain classification features; a normalization module, used to normalize the development cycle days, number of code lines, number of code files, test coverage and number of version defects to obtain numerical features.

[0015] Optionally, the evaluation unit includes: a second acquisition module, used to obtain historical production data and corresponding historical production evaluation results of historical projects before pre-building a target decision tree based on historical production data and historical production evaluation results, and pre-process the historical production data to obtain historical production characteristics; a division module, used to divide the historical production characteristics and the corresponding historical production evaluation results into a training set and a test set.

[0016] Optionally, the evaluation unit also includes: a verification module, which constructs a target decision tree based on the historical production features and historical production evaluation results in the training set, performs a correlation check on each historical production feature and the corresponding historical production evaluation result, and determines the historical production feature whose verification result indicates a first correlation as a leaf node of the decision tree, wherein, when the number of categories of the classified historical production features is greater than a preset classification number threshold, the classified historical production features are subjected to multi-classification feature processing before performing the correlation check; a first determination module, which is used to exclude the historical production features that have been determined as leaf nodes, determine the next leaf node from the remaining historical production features, and iteratively determine the next leaf node until the decision tree meets the preset growth stop condition; a pruning module, which is used to perform post-pruning processing on the decision tree that has stopped growing to obtain the target decision tree.

[0017] Optionally, the verification module includes: a calculation submodule, which is used to calculate the information gain of each category in the classified historical production characteristics when performing multi-classification feature processing on the classified historical production characteristics, and sort the information gains; a combination submodule, which is used to combine the categories corresponding to the first-ranked and second-ranked information gains into target features of the classified historical production characteristics, and use the target features for correlation verification.

[0018] Optionally, the evaluation unit also includes: an input module, which tests the target decision tree constructed by the training set based on the historical production characteristics and historical production evaluation results in the test set, inputs the historical production data into the target decision tree, and obtains the test set production evaluation results and the test set production success probability; an indicator calculation module, which calculates the accuracy, precision and recall of the target decision tree based on the historical production evaluation results and the test set production evaluation results; an adjustment module, which adjusts the parameters of the target decision tree based on the accuracy, precision and recall, and uses the adjusted target decision tree for production risk analysis of the target project.

[0019] Optionally, the calculation unit includes: a sorting module, used to sort the absolute values ​​of the contribution when determining the production risk feature set based on the contribution; a second determination module, used to determine the production feature corresponding to the first-ranked contribution as the first production risk feature; a third determination module, used to determine the production features corresponding to other contribution degrees excluding the first-ranked one as the second production risk feature; and a set generation module, which generates a production risk feature set based on the first production risk feature and the second production risk feature.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is also provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned decision tree-based project commissioning analysis methods.

[0021] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned decision tree-based project commissioning analysis methods.

[0022] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned decision tree-based project commissioning analysis methods.

[0023] In the present disclosure, the production data of the target project is obtained from multiple data sources, the production data is preprocessed to obtain the production characteristics, and then the production characteristics are input into the target decision tree. The target decision tree performs a production risk assessment and outputs the production assessment results and the production success probability. Then, based on the characteristic values ​​of the production characteristics, the contribution of each feature in the production characteristics is calculated, and the production risk feature set is determined based on the contribution. Finally, based on the production assessment results, the production success probability, and the production risk feature set, a production risk analysis report for the target project is generated.

[0024] Based on the above-mentioned public content, a decision tree model is pre-built based on historical production data and historical production evaluation results, and then pruned. The accurate production evaluation results and the probability of successful production are automatically and quickly output according to the input production characteristics. The final generated production risk analysis report can help the project maintenance side identify and evaluate the potential risks of the target project, reduce the uncertainty and potential negative impact of the production of the target project, and thus solve the technical problem of low accuracy in the related technology that relies on manual judgment of the risks of project production. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0026] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a project commissioning analysis method based on a decision tree is shown;

[0027] Figure 2 is a flow chart of an optional project commissioning analysis method based on a decision tree according to an embodiment of the present invention;

[0028] Figure 3 It is a simplified flow chart of an optional project commissioning analysis method based on a decision tree according to an embodiment of the present invention;

[0029] Figure 4 is a schematic diagram of an optional project commissioning analysis device based on a decision tree according to an embodiment of the present invention;

[0030] Figure 5 It is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] It should be noted that the decision tree-based project production analysis and its device in the present disclosure can be used in the field of financial technology. When a project production analysis is performed based on a decision tree, it can also be used in any field other than the field of financial technology. When a project production analysis is performed based on a decision tree, the present disclosure does not limit the application field of the decision tree-based project production analysis method and its device.

[0034] It should be noted that the information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) collected by this disclosure are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data are in compliance with the relevant laws, regulations and standards of the relevant regions, necessary confidentiality measures are taken, and public order and good customs are not violated, and corresponding operation portals are provided for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or organizations. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or organization.

[0035] It should be noted that in the present disclosure, when collecting and analyzing customer information, corresponding operation entrances are provided for users to choose to agree or reject the automated decision results; if the user chooses to reject, the expert decision-making process is entered.

[0036] The following embodiments of the present invention can be applied to various systems / applications / devices for project commissioning analysis based on decision trees. The present invention obtains commissioning data of a target project from multiple data sources, preprocesses the commissioning data, obtains commissioning features, and then inputs the commissioning features into a target decision tree, which performs commissioning risk assessment, automatically and quickly outputs accurate commissioning assessment results and commissioning success probability, and then calculates the contribution of each feature in the commissioning feature based on the feature value of the commissioning feature, determines the commissioning risk feature set based on the contribution, and finally generates a commissioning risk analysis report for the target project based on the commissioning assessment results, commissioning success probability, and commissioning risk feature set, to help the project maintenance end identify and assess the potential risks of the target project and reduce the uncertainty and potential negative impact of the commissioning of the target project.

[0037] The present invention is described in detail below in conjunction with various embodiments.

[0038] Embodiment 1

[0039] According to an embodiment of the present invention, an embodiment of a project commissioning analysis method based on a decision tree is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0040] The project commissioning analysis method embodiment based on a decision tree provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 The hardware structure block diagram of a computer terminal (or mobile device) for implementing a project commissioning analysis method based on a decision tree is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more ( Figure 1 The computer system 102 includes a processor 102 (the processor 102 may include but is not limited to a microprocessor MCU (Microcontroller Unit) or a programmable logic device FPGA (Field Programmable Gate Array)), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a Universal Serial Bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. A person skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0041] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0042] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the project commissioning analysis method based on the decision tree in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, the above-mentioned project commissioning analysis method based on the decision tree is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0043] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0044] The display may be, for example, a touch screen liquid crystal display (LCD), which enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0045] Under the above operating environment, this application provides Figure 2 The project commissioning analysis method based on the decision tree is shown. Figure 2 is a flowchart of an optional project commissioning analysis method based on a decision tree according to an embodiment of the present invention. Figure 2 As shown, the method comprises the following steps:

[0046] Step S201, obtaining the production data of the target project from multiple data sources, preprocessing the production data, and obtaining production features, wherein the production features include: categorical features and numerical features.

[0047] The embodiment of the present invention first needs to obtain the production data of the target project (target software development project) from multiple data sources. These production data can come from project management information systems, financial institution log systems, financial institution version control systems, financial institution testing systems, etc. By obtaining production data from multiple data sources, the comprehensiveness and accuracy of the data can be ensured.

[0048] After obtaining sufficiently comprehensive production data, an embodiment of the present invention preprocesses the production data. The preprocessing process may include data cleaning, duplicate value removal, missing value processing, outlier processing, and data format unification. The production data is converted into production features after preprocessing. High-quality, structured production features can be obtained by preprocessing the production data. Since the obtained production data includes numerical data and categorical data, the production features obtained after preprocessing include categorical features and numerical features. The embodiment of the present invention uses these production features as inputs to the target decision tree model to perform risk assessment before the target project goes into production.

[0049] Step S202, input the production characteristics into the target decision tree, and the target decision tree performs production risk assessment, outputs the production assessment results and the production success probability, wherein the target decision tree is a decision tree model pre-constructed based on historical production data and historical production assessment results, and has undergone pruning processing.

[0050] After obtaining the commissioning characteristics of the target project, the commissioning characteristics are input into the target decision tree, and the target decision tree performs a commissioning risk assessment on the target project. The target decision tree used in the embodiment of the present invention is a trained and optimized decision tree model, which is constructed and trained based on a large amount of historical software development project commissioning data and known commissioning assessment results. In the process of building the model, the decision tree is pruned, and the target decision tree is finally obtained through iterative optimization.

[0051] After inputting the production characteristics of the target project into the target decision tree, the production evaluation results and production success probability of the target project will be obtained. By analyzing the output production evaluation results and production success probability, accurate risk analysis of the target project that is about to be put into production can be carried out to reduce the uncertainty in the production stage.

[0052] Step S203, based on the characteristic value of the production characteristics, calculate the contribution of each characteristic in the production characteristics, and determine the production risk characteristic set based on the contribution.

[0053] When conducting an accurate risk analysis of a target project, determining the characteristic contribution of production characteristics helps identify the characteristics that have the greatest impact on the production evaluation results and the probability of successful production, that is, identifying the characteristic factors with the highest risk.

[0054] The embodiment of the present invention can use SHAP value as the contribution of production feature. SHAP value (full name: SHapleyAdditiveexplanation value) is an indicator used to explain the output results of machine learning models. It can measure the influence (contribution) of each production feature on the output results of the target decision tree. By calculating the SHAP contribution of each production feature, the production risk feature set is determined based on the size and positive and negative of the SHAP contribution. The determined production risk feature set can reflect the characteristic factors with the highest or higher risk of failure in the production of the target project.

[0055] Step S204, generating a production risk analysis report for the target project based on the production assessment results, the production success probability, and the production risk feature set.

[0056] Finally, the embodiment of the present invention generates a production risk analysis report for the target project based on the production evaluation results output by the target decision tree, the production success probability, and the production risk feature set obtained by calculating the SHAP contribution. The production risk analysis report describes the key points of the entire risk analysis process of the target project, points out the main risk points of the target project, and then provides the project maintenance end with a scientific assessment of the risks of the target project, timely discovers the production anomalies of the target project, and repairs them.

[0057] Optionally, the step of preprocessing the production data to obtain production characteristics includes: obtaining all production data, wherein the production data includes at least one of the following: production type, development team level, development cycle days, production team level, requirement complexity, resource availability, code scanning results, test coverage, and number of version defects, wherein the code scanning results include: number of code lines and number of code files; classifying the production type, development team level, production team level, requirement complexity, and resource availability to obtain classified characteristics; normalizing the development cycle days, number of code lines, number of code files, test coverage, and number of version defects to obtain numerical characteristics.

[0058] It should be noted that the production data obtained from multiple data sources in the embodiments of the present invention may include: production type, development team level, development cycle days, production team level, requirement complexity, resource availability, code scanning results (including: number of code lines, number of code files), test coverage (including: unit test coverage, integration test coverage), and number of version defects, among which the production type, development team level, production team level, requirement complexity and resource availability are classified data, and the development cycle days, number of code lines, number of code files, test coverage and number of version defects are numerical data.

[0059] When preprocessing these production data, it is necessary to classify the categorized data and normalize the numerical data.

[0060] When classifying the classified data, it includes: classifying "production type" into "1-service request", "2-business launch", and "3-technical optimization"; classifying "development team level" into "1-high", "2-medium", and "3-low"; classifying "production team level" into "1-high", "2-medium", and "3-low"; classifying "demand complexity" into "1-high", "2-medium", and "3-low"; and classifying "resource availability" into "0-insufficient" and "1-sufficient". After classification, the classified production data is converted into classified production features.

[0061] When normalizing numerical data, we normalize the "development cycle days", "number of code lines", "number of code files", "test coverage", and "number of version defects", and map the values ​​to the [0,1] interval. After normalization, the numerical production data is converted into numerical production features.

[0062] Optionally, before pre-building a target decision tree based on historical production data and historical production evaluation results, the process includes: obtaining historical production data of historical projects and corresponding historical production evaluation results, pre-processing the historical production data to obtain historical production characteristics; and dividing the historical production characteristics and the corresponding historical production evaluation results into a training set and a test set.

[0063] In the embodiment of the present invention, the target decision tree is pre-built and trained based on a large amount of historical software development project commissioning data and known commissioning evaluation results. Before building the model, it is necessary to first obtain a large amount of historical project historical commissioning data and corresponding historical commissioning evaluation results, and pre-process the historical commissioning data to obtain historical commissioning features. It should be noted that the historical commissioning evaluation results are classified features, and their types include "1-failure" and "2-success".

[0064] After preprocessing, the embodiment of the present invention divides the historical commissioning characteristics and the corresponding historical commissioning evaluation results into a training set and a test set. The training set will be used to build the target decision tree, and the test set will be used to train the built target decision tree. For example, the embodiment of the present invention can obtain the commissioning data and corresponding commissioning evaluation results of 300 historical projects in the past, divide 240 groups of data into training sets, and divide the remaining 60 groups of data into test sets.

[0065] Optionally, the step of pre-constructing a target decision tree based on historical production data and historical production evaluation results includes: constructing a target decision tree based on historical production features and historical production evaluation results in a training set, and the step of constructing the target decision tree includes: performing a correlation check on each historical production feature and the corresponding historical production evaluation result, and determining the historical production feature whose verification result indicates a first correlation as a leaf node of the decision tree; wherein, when the number of categories of the classified historical production features is greater than a preset classification number threshold, before performing the correlation check, performing multi-classification feature processing on the classified historical production features; excluding the historical production features that have been determined as leaf nodes, determining the next leaf node from the remaining historical production features, and iteratively determining the next leaf node until the decision tree meets the preset growth stop condition; and performing post-pruning processing on the decision tree that has stopped growing to obtain the target decision tree.

[0066] The embodiment of the present invention uses the training set data of historical projects to construct a target decision tree. In this process, each historical production feature is firstly checked for correlation with the corresponding historical production evaluation result. For the historical production features of different types, the chi-square check strategy can be used to measure the correlation between the historical production features and the historical production evaluation results. For the numerical historical production features, the variance analysis strategy can be used to measure them.

[0067] Among them, the formula for calculating the correlation check value using the chi-square check strategy is:

[0068]

[0069] In the above formula, x 2 is the correlation check value between the calculated historical production characteristics and the corresponding historical production evaluation results, O i is the characteristic value of the historical commissioning characteristics of the i-th historical project, E i is the characteristic value of the historical commissioning evaluation result of the i-th historical project (i.e., the corresponding historical commissioning evaluation result), and k is the number of historical projects. It should be noted that after the correlation check value is calculated, it is compared with the preset critical value. If the correlation check value is greater than the preset critical value, it is determined that there is a significant correlation between the historical commissioning characteristic and the historical commissioning evaluation result.

[0070] In addition, the variance analysis strategy is a statistical strategy that can measure whether the difference between data is significant. When the variance analysis strategy is used to calculate the correlation check value, the correlation check value between the numerical historical production characteristics and the corresponding historical production evaluation results will be obtained. After calculating the correlation check values ​​between all historical production characteristics and the corresponding historical production evaluation results, the historical production characteristics with the first correlation indicated by the check result (i.e., the historical production characteristics with the largest correlation check value) are determined as leaf nodes of the decision tree.

[0071] It should be noted that, since the number of categories of the historical production characteristics of the classification is not uniform, the embodiment of the present invention sets a preset classification number threshold, for example, set to "2". When the number of categories of the historical production characteristics of the classification is greater than 2, multi-classification feature processing will be performed on them before the correlation check is performed. The purpose of performing multi-classification feature processing on the historical production characteristics of the classification with multiple categories is to ensure that the decision tree will not be biased towards the production characteristics with a large number of categories during the splitting process, resulting in overfitting of the decision tree.

[0072] In the process of constructing the target decision tree, the root node is first determined based on the above steps, and then the historical production features that have been determined as the root node are excluded, and the next leaf node is determined from the remaining historical production features. This step is repeated and the next leaf node is iteratively determined until the decision tree meets the preset growth stop condition. The preset growth stop condition can be: the number of samples of the current leaf node is less than 5 samples, all samples of the current leaf node belong to the same category, or the depth of the decision tree reaches 6 layers.

[0073] After the decision tree stops growing, it is post-pruned to obtain the target decision tree. The post-pruning process starts from the bottom layer of the decision tree, and gradually checks each subtree of the decision tree upwards, comparing the error of the current subtree (i.e., the error before pruning) with the error after pruning. If the error after pruning is less than or equal to the error before pruning, the subtree is pruned; if the error after pruning is greater than the error before pruning, the subtree is retained. Repeat this step for the upper leaf nodes until the entire decision tree no longer changes. After the post-pruning process is completed, the decision tree that no longer changes is determined as the target decision tree.

[0074] The embodiment of the present invention constructs and trains a decision tree based on a large amount of historical software R&D project production data and known production evaluation results. In the process of building the model, the decision tree is pruned, and a target decision tree is obtained through iterative optimization. The target decision tree is then used in actual scenarios, which can effectively perform risk analysis before the target project is put into production.

[0075] Optionally, the step of performing multi-classification feature processing on the classified historical production features includes: calculating the information gain of each category in the classified historical production features, and sorting the information gains; combining the categories corresponding to the first-ranked and second-ranked information gains into target features of the classified historical production features, and using the target features for correlation verification.

[0076] When performing multi-classification feature processing on the classified historical production features, the information gain of each category in the classified historical production features is first calculated. The calculation formula is:

[0077]

[0078] In the above formula, A represents the historical production feature, S represents the sample data set of the historical production feature, v represents the classification category of the historical production feature, Values(A) represents all classification categories of the historical production feature, S v represents the sample subset of the classification category v in the sample data set of the historical production characteristics. Entropy(S) is the information entropy of the sample data set of the historical production characteristics. v ) is the information entropy of the sample subset with classification category v, and Gain(S,A) is the information gain of category v.

[0079] After calculating the information gain of each category of the historical commissioning feature, all information gains are sorted, and the categories corresponding to the first and second ranked information gains are combined into the target features of the classified historical commissioning features, that is, the multi-category historical commissioning features are converted into historical commissioning features with 2 categories, and the historical commissioning features with 2 categories are used for correlation verification. Through multi-classification feature processing, the feature space of the decision tree model can be effectively simplified, the overfitting problem of the target decision tree can be prevented, and the efficiency of the target decision tree can be improved.

[0080] Optionally, the step of pre-constructing a target decision tree based on historical production data and historical production evaluation results also includes: testing the target decision tree constructed by the training set based on the historical production features and historical production evaluation results in the test set, and the step of testing the target decision tree includes: inputting the historical production data into the target decision tree to obtain the test set production evaluation results and the test set production success probability; based on the historical production evaluation results and the test set production evaluation results, calculating the accuracy, precision and recall of the target decision tree; based on the accuracy, precision and recall, adjusting the parameters of the target decision tree, and using the adjusted target decision tree for production risk analysis of the target project.

[0081] After the target decision tree is constructed and trained, it is tested using the test set of the historical project. In this process, after the historical production data of the test set is input into the target decision tree, the test set production evaluation results and the test set production success probability will be obtained. Based on the test set production evaluation results and the known historical production evaluation results, the accuracy, precision and recall of the target decision tree are calculated respectively.

[0082] Before calculating the accuracy, precision, and recall, a confusion matrix is ​​first generated based on the test set production evaluation results and the known historical production evaluation results. The confusion matrix is ​​shown in Table 1 below:

[0083] Table 1

[0084] Actual value\predicted value success fail success True Positive(TP) False Negative(FN) fail False Positive (FP) True Negative(TN)

[0085] Among them, "actual value" represents "known historical production evaluation results", and "predicted value" represents "test set production evaluation results". "True Pos it ive (TP)" represents the number of historical projects that are actually successful and predicted to be successful, "False Pos it ive (FP)" represents the number of historical projects that are actually failed but predicted to be successful, "True Negative (TN)" represents the number of historical projects that are actually failed and predicted to be failed, and "False Negative (FN)" represents the number of historical projects that are actually successful but predicted to be failed.

[0086] Based on the above TP, FP, TN and FN, the accuracy, precision and recall are calculated using the following formula:

[0087]

[0088] In the above formula, W1 is the accuracy, W2 is the precision, and W3 is the recall. Finally, based on the accuracy, precision, and recall, the parameters of the target decision tree are adjusted, such as adjusting the depth of the target decision tree or the number of samples of the leaf nodes, and the adjusted target decision tree is used for the production risk analysis of the target project.

[0089] Optionally, the step of determining a set of production risk characteristics based on contribution includes: sorting the absolute values ​​of contribution; determining the production characteristics corresponding to the first-ranked contribution as the first production risk characteristic; determining the production characteristics corresponding to other contribution degrees excluding the first-ranked contribution as the second production risk characteristic; and generating a set of production risk characteristics based on the first production risk characteristic and the second production risk characteristic.

[0090] The embodiment of the present invention calculates the SHAP contribution of each production feature, and then determines the production risk feature set based on the size and positive and negative of the SHAP contribution. It should be noted that the production risk feature set may include: the maximum impact feature and the relatively large impact feature, wherein the maximum impact feature may be the production feature with the largest absolute value of the SHAP contribution, the relatively large impact feature may be several production features other than the production feature with the largest absolute value of the SHAP contribution, and the relatively large impact feature may also be several production features with the largest negative value of the SHAP contribution.

[0091] Figure 3 is a simplified flow chart of an optional project commissioning analysis method based on a decision tree according to an embodiment of the present invention, such as Figure 3As shown, at the beginning of the method execution, historical data is first obtained to build a target decision tree, and the target decision tree is tested and evaluated, and then the target decision tree is used in the actual scenario for the commissioning analysis of the target project. The commissioning data of the target project is input into the target decision tree to obtain the commissioning evaluation results and the high-risk characteristic factors that affect the commissioning evaluation results. Finally, based on the commissioning evaluation results and the high-risk characteristic factors, a commissioning risk analysis report is generated and sent.

[0092] The steps provided by the above-mentioned decision tree-based project commissioning analysis method can obtain the commissioning data of the target project from multiple data sources, pre-process the commissioning data, obtain the commissioning characteristics, and then input the commissioning characteristics into the target decision tree. The target decision tree performs a commissioning risk assessment, outputs the commissioning assessment results and the probability of successful commissioning, and then calculates the contribution of each feature in the commissioning features based on the characteristic values ​​of the commissioning characteristics, determines the production risk feature set based on the contribution, and finally generates a commissioning risk analysis report for the target project based on the production assessment results, the probability of successful commissioning, and the production risk feature set.

[0093] In the embodiment of the present invention, the target decision tree is a decision tree model that is pre-built based on historical production data and historical production evaluation results and has undergone pruning processing. It can automatically and quickly output accurate production evaluation results and production success probability according to the input production characteristics. The final generated production risk analysis report can help the project maintenance end identify and evaluate the potential risks of the target project, reduce the uncertainty and potential negative impact of the target project's production, and thus solve the technical problem of relying on manual judgment of project production risks in related technologies with low accuracy.

[0094] The following is a detailed description in conjunction with another embodiment.

[0095] Embodiment 2

[0096] A decision tree-based project commissioning analysis device provided in this embodiment includes multiple implementation units, each of which corresponds to each implementation step in the above-mentioned embodiment 1.

[0097] Figure 4 is a schematic diagram of an optional project commissioning analysis device based on a decision tree according to an embodiment of the present invention, such as Figure 4 As shown, the project commissioning analysis device based on the decision tree may include: an acquisition unit 41, an evaluation unit 42, a calculation unit 43, and a generation unit 44.

[0098] The acquisition unit 41 is used to acquire the production data of the target project from multiple data sources, pre-process the production data, and obtain the production characteristics, wherein the production characteristics include: classification characteristics and numerical characteristics;

[0099] An evaluation unit 42 is used to input the commissioning characteristics into a target decision tree, and the target decision tree performs a commissioning risk evaluation, and outputs a commissioning evaluation result and a commissioning success probability, wherein the target decision tree is a decision tree model pre-constructed based on historical commissioning data and historical commissioning evaluation results and pruned;

[0100] The calculation unit 43 calculates the contribution of each feature in the production characteristics based on the feature value of the production characteristics, and determines the production risk feature set based on the contribution;

[0101] The generating unit 44 generates a production risk analysis report of the target project based on the production evaluation result, the production success probability and the production risk feature set.

[0102] The above-mentioned project commissioning analysis device based on the decision tree obtains the commissioning data of the target project from multiple data sources through the acquisition unit 41, pre-processes the commissioning data to obtain the commissioning characteristics, and then inputs the commissioning characteristics into the target decision tree through the evaluation unit 42, and the target decision tree performs the commissioning risk evaluation, outputs the commissioning evaluation results and the probability of successful commissioning, and then calculates the contribution of each feature in the commissioning characteristics based on the feature value of the commissioning characteristics through the calculation unit 43, determines the commissioning risk feature set based on the contribution, and finally generates the commissioning risk analysis report of the target project based on the production evaluation results, the probability of successful commissioning, and the set of production risk characteristics through the generation unit 44. The target decision tree in the present disclosure is a decision tree model that is pre-built based on historical production data and historical production evaluation results and has undergone pruning processing. It can automatically and quickly output accurate production evaluation results and production success probability according to the input production characteristics. The final generated production risk analysis report can help the project maintenance end identify and evaluate the potential risks of the target project, reduce the uncertainty and potential negative impact of the target project's production, and thus solve the technical problem of relying on manual judgment of project production risks in related technologies with low accuracy.

[0103] Optionally, the acquisition unit 41 includes: a first acquisition module, used to acquire all production data when pre-processing the production data to obtain production features, wherein the production data includes at least one of the following: production type, development team level, development cycle days, production team level, requirement complexity, resource availability, code scanning results, test coverage, and number of version defects, wherein the code scanning results include: number of code lines and number of code files; a classification module, used to classify the production type, development team level, production team level, requirement complexity and resource availability to obtain classification features; a normalization module, used to normalize the development cycle days, number of code lines, number of code files, test coverage and number of version defects to obtain numerical features.

[0104] Optionally, the evaluation unit 42 includes: a second acquisition module, used to obtain historical production data and corresponding historical production evaluation results of historical projects before pre-building a target decision tree based on historical production data and historical production evaluation results, and pre-process the historical production data to obtain historical production characteristics; a division module, used to divide the historical production characteristics and the corresponding historical production evaluation results into a training set and a test set.

[0105] Optionally, the evaluation unit 42 also includes: a verification module, which constructs a target decision tree based on the historical production features and historical production evaluation results in the training set, performs a correlation check on each historical production feature and the corresponding historical production evaluation result, and determines the historical production feature whose verification result indicates a first correlation as a leaf node of the decision tree, wherein, when the number of categories of the classified historical production features is greater than a preset classification number threshold, the classified historical production features are subjected to multi-classification feature processing before performing the correlation check; a first determination module, which is used to exclude the historical production features that have been determined as leaf nodes, determine the next leaf node from the remaining historical production features, and iteratively determine the next leaf node until the decision tree meets the preset growth stop condition; a pruning module, which is used to perform post-pruning processing on the decision tree that has stopped growing to obtain the target decision tree.

[0106] Optionally, the verification module includes: a calculation submodule, which is used to calculate the information gain of each category in the classified historical production characteristics when performing multi-classification feature processing on the classified historical production characteristics, and sort the information gains; a combination submodule, which is used to combine the categories corresponding to the first-ranked and second-ranked information gains into target features of the classified historical production characteristics, and use the target features for correlation verification.

[0107] Optionally, the evaluation unit 42 also includes: an input module, which tests the target decision tree constructed by the training set based on the historical production characteristics and historical production evaluation results in the test set, inputs the historical production data into the target decision tree, and obtains the test set production evaluation results and the test set production success probability; an indicator calculation module, which calculates the accuracy, precision and recall of the target decision tree based on the historical production evaluation results and the test set production evaluation results; an adjustment module, which adjusts the parameters of the target decision tree based on the accuracy, precision and recall, and uses the adjusted target decision tree for production risk analysis of the target project.

[0108] Optionally, the calculation unit 43 includes: a sorting module, used to sort the absolute values ​​of the contribution when determining the production risk feature set based on the contribution; a second determination module, used to determine the production feature corresponding to the first-ranked contribution as the first production risk feature; a third determination module, used to determine the production features corresponding to other contribution degrees excluding the first-ranked one as the second production risk feature; and a set generation module, which generates a production risk feature set based on the first production risk feature and the second production risk feature.

[0109] The above-mentioned decision tree-based project commissioning analysis device may also include a processor and a memory. The above-mentioned acquisition unit 41, evaluation unit 42, calculation unit 43, generation unit 44, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0110] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the project can be analyzed for production based on the decision tree by adjusting kernel parameters.

[0111] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0112] Embodiment 3

[0113] An embodiment of the present application may provide an electronic device, Figure 5 is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (only one is shown) processor 502, memory 504, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0114] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application, and the processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0115] The processor can call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the production data of the target project from multiple data sources, pre-process the production data, and obtain the production characteristics, wherein the production characteristics include: categorical characteristics and numerical characteristics; input the production characteristics into the target decision tree, and the target decision tree performs production risk assessment, and outputs the production assessment results and the probability of successful production, wherein the target decision tree is a decision tree model that is pre-constructed based on historical production data and historical production assessment results and has undergone pruning; based on the characteristic values ​​of the production characteristics, calculate the contribution of each feature in the production characteristics, and determine the production risk feature set based on the contribution; based on the production assessment results, the probability of successful production, and the production risk feature set, generate a production risk analysis report for the target project.

[0116] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain all production data, wherein the production data includes at least one of the following: production type, development team level, development cycle days, production team level, requirement complexity, resource availability, code scanning results, test coverage, and number of version defects, wherein the code scanning results include: number of code lines, number of code files; classify the production type, development team level, production team level, requirement complexity, and resource availability to obtain classification features; normalize the development cycle days, number of code lines, number of code files, test coverage, and number of version defects to obtain numerical features.

[0117] The processor can also call the information and applications stored in the memory through the transmission device to perform the following steps: obtain historical production data of historical projects and corresponding historical production evaluation results, pre-process the historical production data to obtain historical production characteristics; divide the historical production characteristics and the corresponding historical production evaluation results into training sets and test sets.

[0118] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: construct a target decision tree based on the historical production features and historical production evaluation results in the training set, and the steps of constructing the target decision tree include: performing a correlation check on each historical production feature and the corresponding historical production evaluation result, and determining the historical production feature whose verification result indicates a first correlation as a leaf node of the decision tree; wherein, when the number of categories of the classified historical production features is greater than a preset classification number threshold, before performing the correlation check, performing multi-classification feature processing on the classified historical production features; excluding the historical production features that have been determined as leaf nodes, determining the next leaf node from the remaining historical production features, and iteratively determining the next leaf node until the decision tree meets the preset growth stop condition; performing post-pruning processing on the decision tree that has stopped growing to obtain the target decision tree.

[0119] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: calculate the information gain of each category in the historical production characteristics of the classification, and sort the information gains; combine the categories corresponding to the first and second ranked information gains into the target features of the historical production characteristics of the classification, and use the target features for correlation verification.

[0120] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: testing the target decision tree constructed by the training set based on the historical production characteristics and historical production evaluation results in the test set, and the steps of testing the target decision tree include: inputting the historical production data into the target decision tree to obtain the test set production evaluation results and the test set production success probability; based on the historical production evaluation results and the test set production evaluation results, calculating the accuracy, precision and recall of the target decision tree; based on the accuracy, precision and recall, adjusting the parameters of the target decision tree, and using the adjusted target decision tree for the production risk analysis of the target project.

[0121] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: sort the absolute values ​​of the contributions; determine the production characteristics corresponding to the first-ranked contribution as the first production risk characteristics; determine the production characteristics corresponding to other contribution characteristics excluding the first-ranked contribution as the second production risk characteristics; and generate a production risk characteristic set based on the first production risk characteristics and the second production risk characteristics.

[0122] By adopting the embodiment of the present application, a scheme of a project commissioning analysis method based on a decision tree is provided. The commissioning data of the target project is obtained from multiple data sources, the commissioning data is preprocessed, and commissioning characteristics are obtained, and then the commissioning characteristics are input into the target decision tree, and the target decision tree performs commissioning risk assessment, outputs the commissioning assessment result and the commissioning success probability, and then based on the characteristic value of the commissioning characteristics, the contribution of each feature in the commissioning characteristics is calculated, and the commissioning risk feature set is determined based on the contribution, and finally based on the commissioning assessment result, the commissioning success probability, and the commissioning risk feature set, the commissioning risk analysis report of the target project is generated, thereby achieving the automatic and rapid output of accurate commissioning assessment results and commissioning success probability according to the input commissioning characteristics, helping the project maintenance end to identify and assess the potential risks of the target project, and reducing the uncertainty and potential negative impact of the commissioning of the target project, thereby solving the technical problem of low accuracy in the related technology that relies on manual judgment of the risks of project commissioning.

[0123] It can be understood by those skilled in the art that Figure 5The structure shown is for illustration only, and the electronic device may also be a terminal device such as a smart phone, a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, MID), a PAD, etc. Figure 5 The structure of the electronic device is not limited. Figure 5 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Figure 5 Different configurations shown.

[0124] A person skilled in the art may understand that all or part of the steps in the various decision tree-based project commissioning analysis methods in the above-mentioned embodiments may be completed by instructing the hardware related to the terminal device through a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0125] Embodiment 4

[0126] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the project commissioning analysis method based on the decision tree provided in the first embodiment.

[0127] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the decision tree-based project commissioning analysis methods in the above-mentioned embodiment 1.

[0128] Optionally, in this embodiment, the above storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0129] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the project commissioning analysis method based on the decision tree described in each embodiment of the present application.

[0130] The present application also provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the decision tree-based project commissioning analysis method described in each embodiment of the present application are implemented.

[0131] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0132] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0134] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0135] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0136] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0137] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A project commissioning analysis method based on a decision tree, characterized in that: include: Acquire the commissioning data of the target project from multiple data sources, pre-process the commissioning data, and obtain commissioning features, wherein the commissioning features include: categorical features and numerical features; Input the production characteristics into a target decision tree, and use the target decision tree to perform production risk assessment and output a production assessment result and a production success probability, wherein the target decision tree is a decision tree model pre-constructed based on historical production data and historical production assessment results and pruned; Based on the characteristic values ​​of the production characteristics, calculating the contribution of each characteristic in the production characteristics, and determining the production risk characteristic set based on the contribution; Based on the production evaluation result, the production success probability, and the production risk feature set, a production risk analysis report for the target project is generated.

2. The project commissioning analysis method according to claim 1, characterized in that: The step of preprocessing the production data to obtain production characteristics includes: Acquire all the production data, wherein the production data includes at least one of the following: production type, development team level, development cycle days, production team level, demand complexity, resource availability, code scanning results, test coverage, and number of version defects, wherein the code scanning results include: number of code lines and number of code files; Classify the production type, the development team level, the production team level, the demand complexity and the resource availability to obtain the classification characteristics; The development cycle days, the number of code lines, the number of code files, the test coverage, and the number of version defects are normalized to obtain the numerical features.

3. The project commissioning analysis method according to claim 1, characterized in that: Before pre-building the target decision tree based on the historical production data and the historical production evaluation results, the method includes: Obtain historical production data of historical projects and corresponding historical production evaluation results, pre-process the historical production data, and obtain historical production characteristics; The historical production characteristics and the corresponding historical production evaluation results are divided into a training set and a test set.

4. The project commissioning analysis method according to claim 1, characterized in that: The step of pre-building the target decision tree based on the historical production data and the historical production evaluation results includes: The target decision tree is constructed based on the historical production characteristics and historical production evaluation results in the training set. The steps of constructing the target decision tree include: Performing a correlation check on each of the historical production features and the corresponding historical production evaluation result, and determining the historical production features with a first correlation indicated by the verification result as leaf nodes of the decision tree; Wherein, when the number of categories of the classified historical production features is greater than the preset classification number threshold, before performing the correlation check, multi-classification feature processing is performed on the classified historical production features; Excluding the historical production features that have been determined as leaf nodes, determining the next leaf node from the remaining historical production features, and iteratively determining the next leaf node until the decision tree meets a preset growth stop condition; Post-pruning is performed on the decision tree that has stopped growing to obtain the target decision tree.

5. The project commissioning analysis method according to claim 4 is characterized in that: The step of performing multi-classification feature processing on the historical production features of the classification includes: Calculating the information gain of each category in the historical production characteristics of the classification, and sorting the information gains; The categories corresponding to the first-ranked and second-ranked information gains are combined into target features of the classified historical production features, and the target features are used in the correlation verification.

6. The project commissioning analysis method according to claim 1, characterized in that: The step of pre-building the target decision tree based on the historical production data and the historical production evaluation results also includes: The target decision tree constructed by the training set is tested based on the historical production characteristics and historical production evaluation results in the test set. The step of testing the target decision tree includes: Inputting the historical production data into the target decision tree to obtain the test set production evaluation result and the test set production success probability; Based on the historical production evaluation results and the test set production evaluation results, calculating the accuracy, precision and recall of the target decision tree; Based on the accuracy, the precision and the recall, the parameters of the target decision tree are adjusted, and the adjusted target decision tree is used for production risk analysis of the target project.

7. The project commissioning analysis method according to claim 1, characterized in that: The step of determining a set of production risk characteristics based on the contribution degree includes: sorting the absolute values ​​of the contributions; Determine the production feature corresponding to the contribution degree ranked first as the first production risk feature; Determine the production characteristics corresponding to other contribution degrees excluding the first-ranked one as the second production risk characteristics; The production risk feature set is generated based on the first production risk feature and the second production risk feature.

8. A project commissioning analysis device based on a decision tree, characterized in that: include: An acquisition unit is used to acquire the commissioning data of the target project from multiple data sources, and pre-process the commissioning data to obtain commissioning features, wherein the commissioning features include: categorical features and numerical features; An evaluation unit, used for inputting the production characteristics into a target decision tree, performing production risk evaluation by the target decision tree, and outputting a production evaluation result and a production success probability, wherein the target decision tree is a decision tree model pre-constructed based on historical production data and historical production evaluation results and pruned; A calculation unit, which calculates the contribution of each feature in the production characteristics based on the feature value of the production characteristics, and determines a production risk feature set based on the contribution; A generating unit generates a production risk analysis report of the target project based on the production evaluation result, the production success probability, and the production risk feature set.

9. An electronic device, characterized in that: It includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the decision tree-based project commissioning analysis method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the project commissioning analysis method based on a decision tree as described in any one of claims 1 to 7 are implemented.

Images